GStars
    cactus-compute

    cactus-compute/cactus

    Sub-50mb agentic, voice and vision intelligence for consumer devices.

    llm
    cactus
    edge-ai
    foundation-model
    function-calling
    on-device-ai
    Python
    Apache-2.0
    14.5K stars
    1.0K forks
    81 GitHub watchers
    Updated Nov 7, 2026
    View on GitHub

    Build with Backblaze B2

    SDKs, agent skills, IDE extensions, and reference pipelines from Backblaze Labs. All open source.

    Explore Backblaze Labs

    Loading star history...

    Where it stands

    Rank, growth pace and release cadence relative to everything GStars tracks.

    The project lists 33 contributors on GitHub.

    Peaked at #4 in the weekly ranking the week of Aug 9, its best in the past year.

    Use Cases & Benefits

    • Implements a tiny foundation model enabling on-device automation with 2-bit weights for phones, wearables, smart homes, robots, cars, and microcontrollers.
    • Delivers efficient, low-memory AI capable of tool calling and structured extraction, optimized for resource-constrained edge devices.
    • Use for building lightweight, privacy-preserving AI assistants on mobile and wearable devices that run locally without cloud dependency.
    • Use for integrating function-calling AI models into smart home systems to automate device control with minimal latency and bandwidth.
    • Use for deploying AI-powered automation on microcontrollers and embedded systems in robotics or automotive applications requiring compact models.

    Should you depend on it?

    License terms, project hygiene and risks, from GitHub metadata and GStars' snapshots.

    License

    Apache 2.0

    Permissive: commercial use, modification and redistribution are allowed with attribution; includes an explicit patent grant from contributors.

    Community files

    • README: present

    Watch-outs

    • No GitHub releases; pin a commit or tag instead of a release.

    General guidance from repository metadata, not legal advice. How we score

    From the README

    We distilled Gemini 3.1 into a 26m parameter "Simple Attention Network" that you can even finetune locally on your Mac/PC. In production, Needle runs on Cactus at 6000 toks/sec prefill and 1200 decode speed. Weights are fully open on Cactus-Compute/needle, as well as the dataset generation.

    Needle is an experimental run for Simple Attention Networks, geared at redefining tiny AI for consumer devices (phones, watches, glasses...). So while it beats FunctionGemma-270m, Qwen-0.6B, Graninte-350m, LFM2.5-350m on single-shot function call for personal AI, Those model are have more scope/capacity and excel in conversational settings. Also, small models can be finicky. Please use the UI in the next section to test on your own tools, and finetune accordingly, at the click of a button.

    Opens a web UI at http://127.0.0.1:7860 where you can test and finetune on your own tools. Weights are auto-downloaded.

    Excerpt from the cactus-compute/cactus README on GitHub (Apache-2.0 license). © its authors.Read the full README on GitHub

    Daily snapshots

    Stars gained per day, and how forks and open issues moved, from GStars' daily snapshots

    Aug 12: +383Aug 13: +747Aug 14: +671Aug 15: +483Aug 16: +475Aug 17: +619Aug 18: +372Aug 19: +320Aug 20: +321Aug 27: +137Aug 28: +140Aug 29: +83Aug 30: +98Aug 31: +112Sep 1: +105Sep 2: +107Sep 3: +90Sep 4: +84Sep 5: +71Sep 6: +73Sep 7: +82Sep 8: +86Sep 9: +105Sep 10: +78Sep 11: +73Sep 12: +80Sep 13: +64Sep 14: +27Sep 20: +344Sep 21: +85Sep 25: +103Sep 26: +35

    +10,841 stars over 88 days across 39 snapshot days; best day +747 on Aug 13.

    Forks

    1.0K+726

    Open issues

    6-22

    Compared with peers in LLM

    Fastest-growing repositories GStars tracks in this cohort for the week of Nov 8, by 7-day star growth

    View category →
    cactus-compute/cactus compared with LLM repositories by stars, 7-day star growth, health score and last release
    RepositoryStars+7dHealthLast release
    NVIDIA/auto-ontology11.4K+8.5K75—
    thedotmack/claude-mem115.8K+4.7K75today
    DietrichGebert/ponytail188.4K+4.4K7512 days ago
    affaan-m/ECC300.8K+4.3K7540 days ago
    Graphify-Labs/graphify145.7K+3.8K75today
    cactus-compute/cactus14.5K—31—
    Category median145.7K+4.4K75—
    Compare cactus with auto-ontology, claude-mem, ponytail →

    Discover Repositories

    Search across tracked repositories by name or description